Congestion Avoidance and Load Balancing in Content Placement and Request Redirection for Mobile CDN

With the development of network function virtualization and software-defined network standards, the mobile network operators are interested in integrating content delivery network (CDN) functionalities into the mobile network to enhance their capability for supporting content oriented services. We consider a mobile CDN system, where Base Stations (BSs) are equipped with storage for replicating content. In such a system, BSs cooperation in replying user requests through backhaul links is a widely adopted mechanism. Blindly redirect user requests upon content placement can cause traffic congestion. As a result, congestion avoidance and load balancing is an important issue to be tackled in this scenario. We investigated the joint optimization problem of content placement and request redirection for the BS-based mobile CDN. Specifically, each BS maintains a transmission queue for replying requests issued from other BSs. Network congestion and BSs load balancing can be jointly considered through guaranteeing network stability. We employ the stochastic optimization model to minimize the long-term time-average transmission cost under network stability constraints. By using the Lyapunov optimization technique, we transform the long-term problem into a set of linear programs solved in each short time duration, and we develop an on-line algorithm to efficiently decide content placement and request redirection without requiring a priori knowledge on the random network state information. Through our theoretical analysis, the performance of the algorithm on optimality and network stability is given. The evaluation confirms that our solution can achieve low transmission cost, whilst avoiding congestion and balancing traffic loads.